Workshop Project

The aim of this project is to evaluate the possibility of using T-Coffee along with some structure superimposition algorithm, such as sap, in order to bridge two distantly related groups of sequences. Each group should contain at least one representative structure, and a meaningful structure based alignment of these two structures must exist.

Reasons why T-Coffee's current algorithm is not well suited for that purpose will be discussed, and a new variant algorithm will be proposed.

Establishing how that information can be passed to t_coffee in order to improve a multiple sequence alignmnent quality and the sensitivity of the resulting profile forms the bases of this project.

In a later stage, the possibility of using the t_coffee local consistency measure in order to filter out uninformative residues will be analysed, especialy for its effect on the profile sensitivity/specificity.

Category 3 of BaliBase will be used to provide the testcase, sequences will be analysed and structures retrieved from PDB. Various protocols will be set up using T-Coffee and the quqlity of the resulting alignments will be analysed by comparison with BaliBase, using aln_compare.

The resulting alignment will be turned into a profile using the pftools and used to scan databases.

aln_compare, seq_reformat, sap, process_pdb and t_coffee will be used. They can be downloaded from my home page.